The proposed study investigates the contextual effects of neighborhood socioeconomic status (SES) on the health and well being of older adults. There are four specific aims, to: (1) examine the extent to which the emotional well being of older persons differs, on average, among neighborhoods, and the extent to which these differences are the result of parallel differences in the characteristics of the people who live in these neighborhoods; (2) compare structural and ecological models for multilevel-SES effects on emotional distress among the aged, specifically testing cross-level interactions that operationalize the concept of person-environment fit; (3) explore the extent to which models developed for the impact of SES on emotional distress apply more generally to other aspects of health that are especially relevant to older persons, specifically cognitive impairment and physical illness; and, (4) use the passage of time to sort out, at least in part, the nature of the connections among emotional distress, cognitive impairment, and physical illness, and their joint dependence upon SES among older adults. The proposed study uses existing data from the Asset and Health Dynamics Among Older Adults (AHEAD) study. AHEAD is a longitudinal survey of a nationally representative sample (N = 8,222) of adults aged 70 years and older in 1993, with follow-up data available for 1995 and 1998. This large, representative, and diverse sample means that results have excellent external validity. In addition, 1990 Census data is used to operationalize neighborhood using tract-level data. Census data will be merged with the AHEAD individual-level survey data using geocodes from restricted access cross-reference files. The resultant multilevel data set will contain information on individuals nested within tracts, along with information about how these tracts differ from one another along sociocultural (e.g., ethnic composition), economic (e.g., percent below poverty line), demographic (e.g., percent over age 65), and physical (e.g., percent vacant housing units) dimensions. The primary method of analysis is hierarchical linear regression. The multilevel analysis first estimates the magnitude of between-neighborhood variation in emotional distress, and the extent to which this variation persists net of differences in the social status of the people who live in these different neighborhoods. This residual between-neighborhood variation is then used to test structural and ecological models for the impact of neighborhood-SES, specifically testing the hypothesis that the impact of individual-level SES is conditional upon the SES of one's neighbors.